Communication between individual neurons in neuronal circuits occurs primarily at synapses, specialized structures where receptors for neurotransmitters and neurotrophic factors are localized. Synapses have a very high metabolic demand and are vulnerable to dysfunction and degeneration during aging and in an array of neurological disorders. We have been employing a battery of neuronal cell culture and in vivo models of neurodegenerative disorders in order to identify molecular mechanisms that regulate synaptic plasticity and how abnormalities in synaptic signaling mechanisms may contribute to the pathogenesis of age-related neurodegenerative disorders. --Role of Cell Death Pathways in Synaptic Plasticity: We have found that apoptotic biochemical cascades can be activated locally in synaptic terminals where they may modify synaptic function, mediate structural remodeling of the synapses or induce cell death. Whole-cell patch clamp recordings in cultured rat hippocampal neurons showed that caspase activation in response to apoptotic stimuli selectively decreases AMPA channel activity without decreasing NMDA channel activity. In related studies we have found that caspases play important roles in long-term potentiation of synaptic transmission in the hippocampus suggesting a role for caspases in learning and memory. In addition, we have found that the tumor supressor protein p53 can be activated locally in synapses and can modify mitochondrial functions. Our findings suggest roles for caspases in the modulation of neuronal excitability in physiological settings, and also identify mechanisms whereby caspases and p53 may modify synaptic function. We also found that DNA damage can alter glutamate receptor channel activity by a mechanism involving activation of caspases. Caspase-mediated suppression of AMPA currents may allow neurons with damaged DNA to withdraw their participation in excitatory circuits and undergo apoptosis, thereby avoiding widespread necrosis. These findings have important implications for treatment of patients with cancer and neurodegenerative disorders. --Modification of Synaptic Function by Oxidative Stress: We found that 4-hydroxy-2,3-nonenal (4HN), an aldehyde product of lipid peroxidation, exerts a biphasic effect on NMDA-induced current in cultured rat hippocampal neurons with current being increased during the first 2 h and decreased after 6 h. Similarly, 4HN causes an early increase and a delayed decrease in NMDA-induced elevation of intracellular Ca2+ levels. In contrast, 4HN affects neither the ion current nor the Ca2+ response to alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA). The initial enhancement of NMDA-induced current is associated with increased phosphorylation of the NR1 receptor subunit, whereas the delayed suppression of current is associated with cellular ATP depletion and mitochondrial membrane depolarization. Cell death induced by 4HN is attenuated by an NMDA receptor antagonist, but not by an AMPA receptor antagonist. A secreted form of amyloid precursor protein, previously shown to protect neurons against oxidative and excitotoxic insults, prevented each of the effects of 4HN including the early and late changes in NMDA current, delayed ATP depletion, and cell death. These findings show that the membrane lipid peroxidation product 4HN can modulate NMDA channel activity, suggesting a role for this aldehyde in physiological and pathophysiological responses of neurons to oxidative stress. --Calcium Dysregulation and Synaptic Dysfunction: Most of our early studies of calcium regulation in neuronal plasticity and cell survival were performed in cultured hippocampal neurons with cell death and neurite outgrowth being the end points measured. We have also used a cortical synaptosome preparation to show that amyloid beta-peptide can disrupt synaptic calcium homeostasis by impairing the function of ion-motive ATPases and glucose and glutamate transporters. By evaluating calcium homeostasis in synaptosomes from transgenic mice overexpressing either wild-type of mutant PS1, we provided evidence that PS1 mutations can cause a local disturbance in synaptic calcium regulation that manifests as increased calcium responses to glutamate receptor activation. The latter findings are in agreement with electrophysiological studies of hippocampal slices from wild-type and PS1 mutant mice. If abnormalities in synaptic calcium regulation are critical for the cognitive deficits and neuronal degeneration in AD, then stabilizing synaptic calcium homeostasis should delay the onset or slow the progression of AD. Consistent with a potential therapeutic benefit of agents that stabilize cellular calcium homeostasis, we have shown that agents that inhibit calcium release from ER (dantrolene and xestospongin), block calcium influx through voltage-dependent channels (nifedipine and nimodipine) or buffer intracellular free calcium (BAPTA and calbindin) can protect cultured neurons against the pathogenic actions of PS1 mutations. We are now in a position to determine whether such calcium-based approaches are effective in a mouse model of AD. --Synaptic Energy Metabolism: We tested the hypothesis that PS1 mutations impair synaptic energy metabolism. Cortical synaptosomes from PS1 mutant mice exhibited increased vulnerability to mitochondrial dysfunction and caspase activation in response to exposure to Abeta and metabolic insults compared to synaptosomes from non-transgenic mice and mice overexpressing wild-type PS1. In this case, the alterations in synaptic energy metabolism may be the consequence of a primary effect of PS1 mutations on calcium homeostasis. In other studies we have shown that mitochondrial potassium channels and uncoupling proteins can affect the differentiation and survival of neurons. We are currently elucidating the functions of mitochondrial potassium channels and uncoupling proteins in he regulation of synaptic plasticity.